„This LoRA is trained on high-quality portrait photos to create photorealistic images with natural lighting and skin texture. Works best with prompts like '35mm photo, studio lighting, ultra-detailed, realistic face'. Trained on SDXL. Recommended weight: 0.6–0.8.“

Lana - Polish AI Model

First release of the Zywiec LoRA model – trained on 55 portrait images of the Polish artist Zywiec. The model specializes in realistic female portrait generation with classical elegance and strong facial definition (amber eyes, high cheekbones, calm aesthetic).Base: SDXL 1.0 Resolution: 1024x1024 LoRA Type: Character Recommended weight: 0.6–0.9 Best results with prompts using 'Zywiec' + camera/lighting/style context.

Files with SHA256: f011d7b8c2ac59a97021035167030f1c914fd2b59287fd6faa03d455db349428